A community mining system analyzes objects of different types and relationships between the objects of different types to identify communities. The relationships between the objects have an associated time. The community mining system extracts various features related to objects of a designated type from the relationships between objects of different types that represent the evolution of the features over time. The community mining system collects training data that indicates extracted features associated with members of the communities. The community mining system then classifies an object of the designated type as being within the community based on closeness of the features of the object to the features of the training data.

 
Web www.patentalert.com

< Smart sensor continuously adapting to a data stream in real time using both permanent and temporary knowledge bases to recognize sensor measurements

> Correlation of plant states for feedback control of combustion

> Reverse auction engine for normalizing heterogeneous parameters and scoring proposals

~ 00561